WASHINGTON (AP) – An image circulating online of former President Donald Trump smiling and surrounded by a group of black people doesn't seem unusual at first glance, but upon closer inspection… It's obvious.
Strange lighting and over-perfect details are clues to the fact that they were all generated using artificial intelligence. A photo unrelated to the Trump campaign has surfaced. President Trump is trying to win the support of black voters Polls show who remains loyal to President Joe Biden.
The fake images highlighted in a recent BBC investigation provide further evidence to support warnings that the use of AI-generated images will only increase as the November general election approaches. Experts said it highlighted the danger that all groups, including Latinos, women and older male voters, could be targeted by lifelike images. mislead and confuse At the same time, it demonstrates the need for technology regulation.
In a report released this week, researchers at the nonprofit Center to Combat Digital Hate used several popular AI programs to show how easy it is to create AI. realistic deep fake can do stupid voters. The researchers were able to generate the image below. playing cards Meeting with Russian agents biden Many of these AI programs are stuffing ballot boxes at polling places or enlisting armed militias, even though they say they have rules against this type of content.
The center analyzed some of the recent deepfakes of President Trump and Black voters, at least one of which was originally created as satire but is now being used as evidence of Trump's support among Black voters. was determined to be shared by.
social media platform and AI companies I need to do more Imran Ahmed, CEO and founder of the center, said this was to protect users from the harmful effects of AI.
“If a picture is worth a thousand words, these dangerously susceptible image-generating tools, combined with the shady content moderation efforts of mainstream social media, have exposed the bad guys to a level never seen before. has become a powerful tool to mislead voters,” Ahmed said. . “This is a wake-up call for AI companies, social media platforms, and lawmakers. If we don't act now, we're putting American democracy at risk.”
The images raised alarm on both the right and the left, saying they could mislead people into believing the former president was popular among African Americans. Some Trump supporters have expressed frustration with the circulation of the fake images, believing the fabricated scenes will hurt the Republican Party's outreach to black voters.
“This would not have happened if the photo of President Trump and a black person had not been posted on his official campaign and surrogate pages,” said Deante Johnson, president of the Black Conservative League. “It's nonsense to think that the Trump campaign has to use AI to show support for Black people.”
Experts say AI-generated deepfakes can help campaigns attract or mislead specific voters in key battleground states, such as Latinos, women, Asian Americans, and older conservatives. , we hope to see further efforts to target other demographics that they hope to scare.and Dozens of countries hold elections Here are the challenges posed by deepfakes this year: global problem.
In January, voters in New Hampshire received robocalls imitating Mr. Biden's voice, falsely telling them that if they voted in the state's primary, they were ineligible to vote in the general election.later became a political consultant Authorized to make robocallThis may be the first known attempt to use AI to interfere with US elections.
Such content, even if unbelievable, can have harmful effects. According to a February survey It is by researchers at Stanford University and is investigating the potential impact of AI on Black communities. When people realize that the images they see online are unreliable, they may begin to think: Discount legitimate sources.
“As AI-generated content becomes more prevalent and difficult to distinguish from human-generated content, individuals may become more skeptical and distrustful of the information they receive,” the study says. they wrote.
Even if it does not succeed in deceiving a large number of voters, AI-generated content When it comes to voting, candidates, and elections, it can be difficult for anyone to separate fact from fiction, resulting in the marginalization of legitimate sources of information, promoting a loss of trust, and undermining democracy. trust is eroded and political polarization increases.
False claims about candidates and elections are nothing new, but thanks to AI, creating lifelike images, video, and audio is now faster, cheaper, and easier than ever before. . AI deepfakes can reach millions of people before they are published on social media platforms like TikTok, Facebook, and X. high tech companiesgovernment officials and legitimate news outlets are also aware of its existence.
“AI has simply accelerated the misinformation and accelerated it,” said Joe Paul, an executive and advocate who has worked to increase digital access among communities of color. Paul said black communities often have a “history of distrust” of major institutions, including politics and the media, which makes them less aware of public discourse about them and fact-checking to inform their communities. He pointed out that he was becoming more skeptical about the
Digital literacy and critical thinking skills are one of the defenses against AI-generated misinformation, Paul said. “The goal is to enable people to critically evaluate the information they encounter online. The ability to think critically is a lost skill not just in the Black community, but in all communities.”